Perhaps the salient difference between Eli's and Ronald's approach, and that of other cult researchers, is that they're Public Choice economists, so by and large they look at the behavior of people in such groups as "rational" at their core. If people change their beliefs it's in order to realize a rational desire, even if focus of that desire is not, itself, rational. (I know, we don't change our non-rational beliefs for rational reasons, do we?)

I haven't found Berman's paper yet, but you can downlead Wintrobe's here: Can Suicide Bombing Be Rational? (He's a Canadian, it turns out.) And you can read the introduction to his upcoming book here: Rational Extremism. You'll need a certain tolerance for economics jargon, but here's the conclusion from his paper:

Conclusion
In this chapter I developed a simple model to explain how it is possible for a person to rationally commit suicide to further the goals of a group. In the model, an individual gives up autonomy for solidarity, that is he trades his beliefs for a feeling of belonging-ness to a group. Small trades of this type do not result in unusual behaviour and indeed, most of us engage in such behavior all of our lives. However, at large levels, such trades imply that a person is more and more giving up his identity for that of the group, perhaps as personified by its leader, and losing the capacity to make decisions based on values other than those of the leader. Consequently, the choice of larger levels of solidarity may drive a person close to or at a corner solution where her values are entirely those of the leader. Such a person is capable of rational suicide for the goals of the group. Some implications of this view are that small price effects will not change the behaviour of the individual in question, and even very fairly large ones might not cause the person to revert to her old identity since he has given it up in exchange for solidarity [note the distracting Canadian-induced use of politically correct gender references]. However, very large changes will cause a very substantial change, as is typical in the analysis of corner solutions. ["Corner solution" is a reference to a micro-economics concept where an individual has relinquished all of his assets to buy a particular good.]

Although such slavish devotion to the group is typically associated with cults, modern terrorist groups seem capable of producing such individuals even though they are relatively large and dispersed. The structure of Al Qaeda is not only hierarchical but cellular, and this facilitates vertical control under these circumstances.

Eli's paper takes a somewhat different route than Wintrobe's in some respects, and I think he has a more coherent mechanism for the transition of the group, as well as some concrete recommendations (such as, never "subsidize" a religious group, because it only causes it to become more radical). Eli notes the gradually increasing radicalization of he Jewish Orthodox groups as a result of their ability to demand subsidies in exchange for joining a party to form a government. But the same mechanism holds for Islamist groups. If the religious Islamist groups in Europe ever become powerful enough politically to demand subsidies, Europe is in deep deep trouble.